Evolutionary Algorithms and Simulation for Intelligent Autonomous Vehicles in Container Terminals

The study of applying soft computing techniques, such as evolutionary computation and simulation, to the deployment of intelligent autonomous vehicles (IAVs) in container terminals is the focus of this thesis. IAVs are a new type of intelligent vehicles designed for transportation of containers in container terminals. This thesis for the first time investigates how IAVs can be effectively accommodated in container terminals and how much the performance of container terminals can be improved when IAVs are being used. In an attempt to answer the above research questions, the thesis makes the following contributions: First, the thesis studies the fleet sizing problem in container terminals, an important design problem in container terminals. The contributions include proposing a novel evolutionary algorithm (with superior results to the state-of-the-art CPLEX solver), combining the proposed evolutionary algorithm with Monte Carlo simulation to take into account uncertainties, validating results of the uncertain case with a high fidelity simulation, proposing different robustness measures, comparing different robust solutions and proposing a dynamic sampling technique to improve the performance of the proposed evolutionary algorithm. Second, the thesis studies the impact of IAVs on container terminals’ performance and total cost, which are very important criteria in port equipment. The contributions include developing simulation models using realistic data (it is for the first time that the impact of IAVs on containers terminals is investigated using simulation models) and applying a cost model to the results of the simulation to estimate and compare the total cost of the case study with IAVs against existing trucks. Third, the thesis proposes a new framework for the simulations of container terminals. The contributions include developing a flexible simulation framework, providing a user library for users to create 3D simulation models using drag-and-drop features, and allowing users to easily incorporate their optimisation algorithms into their simulations.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

Liverpool John Moores University

Year dc:date.issued

2015

Author dc:creator

• Kavakeb, S

Contributors dc:contributor

• Nguyen, TT
• Yang, Z

• TC Hydraulic engineering. Ocean engineering

DOI dc:identifier.doi

https://doi.org/10.24377/LJMU.t.00004380

OAI identifier oai:identifier

oai:researchonline.ljmu.ac.uk:4380